Selection Bias in Colorectal Surgery in a Non-Tertiary Hospital: Laparoscopic Versus Open Surgery

Abstract

Introduction:

Laparoscopy is used increasingly to treat malignant and benign colorectal surgical diseases. However, this practice is still not offered to all patients. Many barriers halt the widespread use of laparoscopic colorectal surgery. Both surgeon's and patient's factors contribute to limit a wider use of laparoscopy in colorectal surgery.

Materials and Methods:

We retrospectively analyzed 408 consecutive colorectal resections in a 4-year period, to find out if a selection bias exists in using laparotomy or laparoscopy for colorectal surgical diseases, and which factors are associated with a poor use of laparoscopy or to a preferred laparotomy.

Results:

In our practice, advanced disease, American Society of Anesthesiologist class III and IV, and emergency status are all patient-related factors associated with laparotomy. Surgeon's age more than 52 years and lack of laparoscopic training are surgeon-related factors that negatively affect the chance of being operated on with the laparoscopic technique.

Conclusions:

An extensive laparoscopic colorectal training and a supporting environment, especially during the night shift, are needed to facilitate the use of laparoscopy in colorectal surgery avoiding a bias in selecting surgical candidates to one technique or another.

Introduction

Laparoscopy is used increasingly to treat both benign and malignant surgical colorectal diseases.^1–3 Recent data suggest that laparoscopic colorectal surgery (LCS) is safe and can provide the neoplastic patients with the same short- and long-term results when compared with open surgery.^4–6 The clinical outcome of surgical therapy trial demonstrated similar disease-free survival in patients who underwent resection by laparotomy or laparoscopy to treat colon cancer.⁷ Moreover, especially true for patients with benign diseases, laparoscopy is associated with a shorter hospital stay, earlier return to normal activity, and improved cosmetic results.^8,9

Indications to LCS are the same as the open surgery and even though some controversy exists on the use of laparoscopy for rectal cancer, only few clinical situations (e.g., diffuse fecal peritonitis, massive intestinal obstructions) are clear indications to open surgery. According to recent reports,¹⁰ LCS should be performed in up to 90% of colorectal surgeries with a conversion rate around 10%. However, LCS is not performed universally¹¹ and, most importantly, in the “real world” of current practice, indications to the laparoscopic approach are usually based on patient's clinical conditions, age, American Society of Anesthesiologists (ASA) score, tumor stage, emergency versus elective, and surgeon's skills and training.^12–15

Patients with severe comorbidities, advanced disease, older than 80 years, with ASA class III or higher are usually treated with a planned laparotomy. Surgeon's preference (i.e., skills and experience) may also play an important role in choosing the surgical procedure.¹³ We also believe that a supporting staff in the operating room could facilitate surgeons during the learning curve.

Our institution is a nontertiary hospital and provides a 24-hour emergency surgical service, and elective surgical oncology cases with about 1600 major and minor surgical procedures performed each year. Colorectal surgeries are about 100 each year. Staff surgeons are graduates from a General Surgery Residency Program and half of the personnel completed training in minimally invasive colorectal surgery. All the surgeons have large experience in both elective and emergency colorectal cases. In the emergency, the on call surgeon determines the treatment (laparoscopy versus laparotomy) based on clinical parameters, radiology, and personal experience. All the elective cases are discussed in a multidisciplinary meeting. In the elective cases, an effort is made to perform all the procedures through laparoscopy; however, the assigned surgeon decides the technique according to clinical and radiological assessment and personal skills.

Clinical charts of patients who underwent colorectal surgery at our institution were retrospectively analyzed comparing laparoscopic versus open surgery with the intent of defining the surgeon's decision process in choosing laparoscopic technique in our practice.

We hypothesized that laparoscopic-trained surgeons would perform colorectal resection by minimally invasive surgery more often. Moreover, a helpful and “friendly” environment (such as daily shift, elective versus emergency) would allow a plan for laparoscopic surgery creating a bias in choosing the surgical modality (laparoscopy versus laparotomy). We also evaluated the site of the tumor to find out if different segments of the colonic tract were treated differently.

Patients and Methods

We retrospectively review all patients who underwent colorectal surgery (both emergency and elective) at out institution in a 4-year period, from October 2011 to December 2015. Four hundred eight patients underwent colorectal surgery for both benign and malignant disease. All the attending surgeons were involved in the surgeries. We analyzed patient's and surgeon's variables and matched these variables with the selected surgical technique. If the patient was converted to open surgery, the case was assigned to the open group. This assumption means that a low threshold to convert is considered not offering the patients laparoscopy at all.

Patients' variables

Patients were classified into two main groups (group 1: laparoscopy LS; group 2: open surgery OS). Patient's variables such as age, sex, ASA score, site of colonic resection, benign or malignant disease, emergency and elective surgery have been evaluated and matched with the surgical technique. Due to technical difficulty in treating different segments of the colorectal tract, we arbitrarily classified the patients into the following categories: group 1: right and left colon; group 2: transverse colon and right and left flexures; and group 3: rectum and anus. We assumed that these groups of patients represented for the surgeons the same level of technical challenge. Patients with neoplastic disease have been also evaluated for tumor extension. Tumors invading adjacent organs or the abdominal wall, tumor with metastases to distant organs, or synchronous colorectal neoplasms were all considered “macroscopically advanced diseases.” Adhesions and/or prior surgery were not included among patient's variables.

Surgeons' variables

Surgeons were classified into two groups (1: minimally invasive trained [MIT], and 2: without training in minimally invasive surgery [NMIT]). Surgeon's age was also matched with the surgical technique and we dichotomized age in two categories based on median age (≤52 years, >52 years). The age of the surgeon was fixed at the year 2011 when the analysis started. Variables such as night or day shift have been also included in the surgeon's variables and matched with the surgery.

Statistical analysis

We calculated mean and standard deviation for continuous variables and frequency for categorical variables.

Two-tailed chi-square test was used to identify associations between covariates and laparotomy technique. The covariates associated were entered in a multiple logistic regression model. The odds ratios (OR) and the 95% confidence intervals (95% CI) were calculated. The significant level was set at 0.05. All analyses were done by using Stata MP/12 software (@ StataCorp LP).

Results

One hundred seventy-five patients (43%) underwent laparoscopic surgery and 233 (57%) underwent open surgery. Overall, ASA score was I–II in 183 patients (45%), and ASA score was III–IV in 225 patients (55%) (Table 1).

Table 1.

Patients' Characteristics (n = 408)

	Mean ± SD or n (%)
Age	72 ± 13.7
Sex
Male	213 (52.2)
Female	195 (47.8)
Surgical technique
Laparoscopy	175 (42.9)
Laparotomy	233 (57.1)
ASA score
I	18 (4.4)
II	165 (40.4)
III	197 (48.3)
IV	28 (6.9)
Diagnosis
Right colon tumor	102 (25)
Transverse colon tumor	22 (5.4)
Left colon tumor	32 (7.9)
Sigmoid tumor	72 (17.7)
Rectal tumor	63 (15.4)
Diverticulitis	49 (12)
Intestinal ischemia	9 (2.2)
Anal tumor	1 (0.2)
Miscellaneous	58 (14.2)
Site of disease
I (right and left colon)	268 (65.69)
II (right and left colic flexures and transverse colon)	52 (12.74)
III (rectus and anus)	82 (20.1.3)
Advanced disease
Yes	162 (39.7)
No	243 (60.3)
Emergency
Yes	126 (30.6)
No	283 (69.4)

ASA, American Society of Anesthesiologists; SD, standard deviation.

Two hundred forty-six (60%) patients out of 408 were treated by surgeons aged ≤52 years (Table 2).

Table 2.

Surgeons' Variables (n = 408)

	n (%)
Surgeon age
≤52	246 (60.3)
>52	162 (39.7)
Laparoscopic training
Yes (MIT)	213 (52.2)
No (NMIT)	195 (47.8)
Time of surgery
Day	175 (42.9)
Night	233 (57.1)

MIT, minimally invasive trained; NMIT, without training in minimally invasive surgery.

Results of the univariate analysis for factors associated with laparotomy are listed in Table 3. No significant differences were found between patients treated with laparotomy and patients treated with laparoscopy, with regard to gender and site of disease.

Table 3.

Patient's and Surgeon's Variables Associated with Laparotomy (Covariate with Crude Odds Ratios)

Variables, surgical technique
	Laparoscopy, n = 175, n (%)	Laparotomy, n = 233, n (%)	χ²	P	OR	95% CI
Patient's age
<65	82 (47.40)	91 (52.60)	50.42	<.001	Ref.
65–79	67 (63.81)	38 (36.19)			0.51	0.31–0.84
>79	26 (20.00)	104 (80.00)			3.60	2.14–6.08
	82 (47.40)	91 (52.60)
Sex
F	77 (39.49)	118 (60.51)			Ref.
M	98 (46.01)	115 (53.99)	1.77	.184	0.77	0.52–1.13
ASA score
I + II	117 (63.93)	66 (36.07)	59.98	<.001	Ref.
III+IV	58 (25.78)	167 (74.22)			5.10	3.34–7.80^a
Emergency
No	151 (53.36)	132 (46.64)	41.30	<.001	Ref.
Yes	24 (19.20)	101 (80.80)			4.81	2.91–7.96^a
Night shift
No	168 (45.04)	205 (54.96)	8.19	.003	Ref.
Yes	7 (20.00)	28 (80.00)			3.28	1.40–7.69^a
Surgeon's age
≤52	127 (51.63)	119 (48.37)	19.29	<.001	Ref.
>52	48 (29.63)	114 (70.37)			2.53	1.67–3.86^a
Training
No	17 (14.91)	97 (85.09)			Ref.
Yes	158 (53.74)	136 (46.26)	50.56	<.001	0.15	0.09–0.27^b
Advanced disease
No	125 (50.81)	121 (49.19)	16.14	<.001	Ref.
Yes	50 (30.86)	112 (69.14)			2.31	1.53–3.51^b
Site of disease
I	116 (43.28)	152 (56.72)	5.90	.052	Ref.
II	15 (28.85)	37 (71.15)			1.88	0.98–3.59
III	41 (50.00)	41 (50.00)			0.76	0.46–1.25

CI, confidence intervals; OR, odds ratios.

Only 29% of patients with tumor located in the right or left flexure underwent laparoscopy.

The patient's covariates associated with laparotomy included the following: older age (OR for age >79 versus <65 years, 3.6, 95% CI: 2.14–6.08), higher ASA score, emergency status, and advanced disease (69% of patients with tumor invading adjacent organs, distant organ metastases, and synchronous tumors had laparotomy).

With respect to surgeon's variables, the significant factors associated with laparotomy were as follows: night shift and surgeon's age >52 years. MIT surgeons operated on 54% of the patients by laparoscopy, while they operated on only 46% of the cases by laparotomy. NMIT group of surgeons performed 97 of the procedures by laparotomy (85%) and they operated on only 17 patients by laparoscopy. Previous training, as reported in Table 3, was a protective factor for laparotomy.

According to the multivariate logistic regression model, the significant predictors for laparotomy were as follows: higher ASA score, emergency status, advanced disease, surgeon's age >52 years. A formal training in minimally invasive surgery was, on the contrary, a factor that protected the patients from laparotomy, as reported in Table 4.

Table 4.

Patient's and Surgeon's Variables Associated with Laparotomy (Multivariate Regression)

Variables	OR	P	95% CI
ASA score
I + II	Ref.
III+IV	4.91	.000	3.01–8.00
Emergency
No	Ref.
Yes	3.93	.000	2.19–7.06
Advanced disease
No	Ref.
Yes	2.42	.001	1.46–4.00
Surgeon's age
≤52	Ref.
>52	1.04	.019	1.01–1.07
Minimally invasive training
No	Ref.
Yes	0.16	.000	0.08–0.30

Table 5 shows differences in terms of length of stay and complications between the two groups of patients.

Table 5.

Postoperative Outcomes

	Open (233)	Laparoscopic (175)
Complications	66 (28.52%)	49 (28%)
Event ratio	4 (1.71%)	0
Anastomotic leak	17 (7.29%)	16 (9.14%)
Intestinal ischemia	1 (0.43%)	2 (1.14%)
Surgical site infection	12 (5.15)	14 (8%)
Incisional hernia	2 (0.85%)	1 (0.57%)
Lesion to the ureter	1 (0.43%)	3 (1.71%)
Obstruction	5 (2.14%)	3 (1.71%)
Bowel perforation	5 (2.14%)	4 (2.28%)
Bleeding	6 (2.57%)	3 (1.71%)
Death	15 (6.43%)	5 (2.85%)
Miscellaneous	5 (2.14%)	3 (1.71%)
Length of stay	15 [3–90]	9 [4–49]

Discussion

The beginning of LCS dates back to 1991 when Jacobs reported on his first 20 patients undergoing laparoscopic-assisted colectomy.¹⁶ He stated that, “although laparoscope-assisted colonic surgery may still be considered a procedure in evolution, we feel that in time it has the potential to be as popular as laparoscopic cholecystectomy.” Unfortunately, LCS has not yet gained the same widespread use as laparoscopic cholecystectomy. This is, in part, due to the initial reports, questioning the safety and, most importantly, the efficacy of this procedure for oncological cases.^17–19 More recently, many reports^6,7,20,21 dissolved all the doubts about the oncologic efficacy for colorectal cancer treated by laparoscopy. Nonetheless, many other factors halt the widespread use of this procedure. Some appear to be surgeon related; others are related to patient and tumor features.

LCS is a complex procedure that implies a longer operative time and a steeper learning curve.^22,23 Consequently, many surgeons are still reluctant to adopt this technique as the technique of choice, and a selection bias plays a role in deciding whether to perform a case by laparoscopy or by laparotomy. As a matter of fact, worldwide, LCS is performed only in a few centers and only in 20 to 35% of all the cases.^22,24,25 Significant predictors of offering a laparoscopy are recent graduation, male sex, location of practice, academic affiliation, and minimally invasive training.^13,15 Lack of available operative time and lack of formal training are, on the contrary, predictors of laparotomy. In another report,²⁶ insurance status, geographic location, and hospital type played a significant role in the selection of a laparoscopic colorectal resection. In other words, diffuse implementation of LCS is still poor. Other factors rather than safety and efficacy play a role. Technical difficulties represent an important obstacle for the operating surgeon. Tumor size and tumor location might render the procedure more difficult. A less advanced T stage, a smaller tumor size, and tumor in the right colon were associated with more use of laparoscopy.^13,15

Our data found no difference between the open and laparoscopic technique as far as morbidity and mortality are concerned. Data differed significantly between the two groups when demographics or clinical factors are analyzed. Variables such as patient's age, ASA status, tumor stage, emergency, and site of colonic resection are all factors that in our practice influenced surgeon's decision in choosing the surgical technique to perform colorectal resections. Clearly, laparoscopy was the method of choice in elective cases to treat younger patients with benign conditions, with an ASA score of I or II, or with a low tumor stage. Even though we found no difference in treating benign or malignant disease by laparoscopy, there is a tendency to treat benign conditions by laparoscopy.

We also noticed a tendency to treat by laparotomy more challenging cases, such as tumors located at the right or left flexures. Again, there was no significant difference in our series, but only 29% of patients with tumor located in the right or left flexure underwent laparoscopy.

Among general surgeons, without fellowship in colorectal surgery, there is a tendency to deny laparoscopy to “high-risk” patients.¹⁴ Excluding high-risk patients from a laparoscopic approach does not imply better results or safer procedures. As a matter of fact, the so-called high-risk patients could potentially benefit the most from a laparoscopic approach.¹⁴ In our practice, a higher ASA level (III or IV) and an emergency were predictive factors of laparotomy.

As far as surgeon-related factors are concerned, age and previous training made the difference at our institution. Almost all the NMIT surgeons performed the procedures by laparotomy, but it needs to be noticed that, even in the MIT group, not all the procedures were started with the laparoscope. In about half of the cases, the minimally invasive trained surgeon performed a laparotomy. When a case was converted, it was included in the open group due to the assumption that a low threshold to convert is the same as not offering laparoscopy at all.

In our division, 4 out of 8 surgeons completed training in either minimally invasive or colorectal surgery. Undoubtedly, surgeon's personal skills and training contributed largely to the technical choice. Our general surgery practice reflects that of many hospitals in our country, and clearly, the surgeon's choice is based on factors that are not strictly clinical or patient related.

We “labeled” night and day shift as a surgeon-related factor. We strongly believe that at night, a strong individual experience can surpass the overwhelming environmental factors that ask for an easier and quicker operation (namely laparotomy). Few patients were operated on by night. Of these, only 7of them received a laparoscopy technique. However, the odds ratio of receiving a laparotomy at night was 3.28.

Factors potentially associated with a better outcome or an easier operation favored the choice of laparoscopy. Tumor stage and site of colonic resection represent, on the contrary, challenging issues for the general surgeons and oriented the choice to perform open or laparoscopic surgery. Usually, tumor located on the transverse colon or close to the flexures challenges the surgeon. T3 or more advanced stages could potentially include a riskier operation.

Environmental factors, namely OR, personnel training, anesthesiologist skills, senior supervision, could potentially interfere with the choice in emergency and nighttime scenarios. Surgeons operating during the night shift in an emergency setting could have been concerned about the operative time. In looking at ways to improve these results in the future, one could ask whether emergency patients should be referred to laparoscopic- or colorectal-trained surgeons even during the nighttime. Interestingly enough, years of practice do not translate into a wider use of laparoscopy. On the contrary, in the Canadian survey,¹³ the surgeons performing laparoscopy were younger and had fewer years of practice. The number of years in clinical practice means older age and consequently is associated with laparotomy. In our practice, 52 years was the surgeon's age cutoff to prefer laparoscopy. Only recently, trained surgeons feel comfortable in performing laparoscopic colorectal resections. This could be related to a major exposure to laparoscopy during the residency training program or to fellowship programs in laparoscopy or simply by a different perception of minimally invasive surgery. Older surgeons should not refuse the laparoscopic approach with the idea that age is an obstacle to climb the learning curve. In a report from China (26), two groups of surgeons (aged 47–54 and 34–43 years, respectively) were compared as far as the short-term outcome in LCS was concerned. They concluded that older surgeons could master the laparoscopic skills easier and faster than the younger ones.

Lack of operating time is one of the reported reasons to avoid laparoscopy.¹³ Busy scheduling in the operating room will always be present, especially in a nontertiary center in an era of financial constrain. This factor could not be easily modified. However, one must consider that laparoscopy, when performed by an experienced hand, could potentially reduce operative time and costs.

Mentorship of an experienced minimally invasive surgeon could be the key to increase the number of cases performed laparoscopically, rather than weekend courses.

Our study is representative of contemporary management of a relatively common surgical disease by a multispecialty group. Surgeon's characteristics are a potent predictor of the operation performed and even though colorectal procedures are complex and require a steep learning curve, surgeons should make every effort to offer this procedure to all the patients, recent literature being clear about the safety and efficacy of this procedure in oncological and benign conditions. Laparoscopy is not inferior to open surgery as far as short- and long-term outcomes are concerned and is superior with respect to length of stay (more than 2 days in our series), complications, and cosmetic results. All the factors affecting the selection bias in colorectal surgery should be avoided and institutions should invest resources in training laparoscopic surgeons to increase the number of minimally invasive procedures.

Conclusion

In our institution, the laparoscopic technique was used more frequently in elective cases, in patients with a younger age, lower ASA status, and a lower tumor stage. Not surprisingly, the open approach was also preferred for patients with more challenging resections such as transverse or right and left colonic flexure. Morbidity and mortality, as well as a long-term survival in patients with tumor, did not differ between the two groups. Due to the safety and the excellent results achieved with LCS as reported in the literature, an effort should be made by general surgeons to increase the use of laparoscopic technique. An extensive laparoscopic colorectal training and a supporting environment, especially during the night shift or in the emergency setting, are needed to facilitate the use of laparoscopy in colorectal surgery avoiding a bias in selecting surgical candidates to one technique or another.

Footnotes

Disclosure Statement

No competing financial interests exist.

References

Moussa

, Rajaganeshan

, et al. Single-center comparative study of laparoscopy versus open colorectal surgery: A 2-year experience. Surg Laparosc Endosc Percutan Tech, 2012; 22:29–32.

Biondi

, Grosso

, et al. Laparoscopic vs open approach for colorectal cancer: Evolution over time of minimal invasive surgery. BMC Surg, 2013; 13:S12.

Chand

, Siddiqui

, et al. Systematic review of emergent laparoscopic colorectal surgery for benign and malignant disease. World J Gastroenterol, 2014; 20:16956–16963.

Bai

, Chen

, et al. Five-year long-term outcomes of laparoscopic surgery for colon cancer. World J Gastroenterol, 2010; 6:4992–4997.

Lacy

, Garcia-Valdecasas

, et al. Laparoscopy-assisted colectomy versus open colectomy for treatment of non-metastatic colon cancer: A randomized trial. Lancet, 2002; 359:2224–2229.

Lacy

, Delgado

, et al. The long term results of a randomized clinical trial of laparoscopic-assisted versus open surgery for colon cancer. Ann Surg, 2008; 248:1–7.

Nelson

, Sargent

, et al. A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med, 2004; 350:2050–2059.

Vlug

, Wind

, et al. Laparoscopy in combination with fast track multimodal management is the best perioperative strategy in patients undergoing colonic surgery: A randomized clinical trial (LAFA study). Ann Surg, 2011; 254:868–875.

Nishikawa

, Ishihara

, et al. Short-term outcomes of open versus laparoscopic surgery in elderly patients with colorectalcancer. Surg Endosc, 2016; 30:5550–5557.

10.

Newman

, Arnold

, et al. The majority of colorectal resections require an open approach, even in units with a special interest in laparoscopic surgery. Colorectal Dis, 2012; 14:29–34.

11.

Mavrantonis

, Wexner

, et al. Current attitudes in laparoscopic colorectal surgery. Surg Endosc, 2002; 16:1152–1157.

12.

Guend

, Lee

, et al. Technique of last resort: Characteristics of patients undergoing open surgery in the laparoscopic era. Surg Endosc, 2015; 29:2763–2769.

13.

Moloo

, Haggar

, et al. The adoption of laparoscopic colorectal surgery: A national survey of general surgeons. Can J Surg, 2009; 52:455–462.

14.

Marks

, Kawun

, et al. Redefining contraindications to laparoscopic colorectal resection for high-risk patients. Surg Endosc, 2008; 22:1899–1904.

15.

Stormark

, Soreide

, et al. Nationwide implementation of laparoscopic surgery for colon cancer: Short-term outcomes and long-term survival in a population-based cohort. Surg Endosc, 2016; 30:4853–4864.

16.

Jacobs

, Verdeja

, et al. Minimally invasive colon resection (laparoscopic colectomy). Surg Laparosco Endosc, 1991; 1:144–150.

17.

Wexner

, Cohen

. Port site metastases after laparoscopic surgery for cure of malignancy. Br J Surg, 1995; 82:295–298.

18.

Berends

, Kazeimer

, et al. Subcutaneous metastases after laparoscopic colectomy. Lancet, 1994; 344:58.

19.

Peters

, Bartles

. Minimally invasive colectomy: Are the potential benefits realized?. Dis Col Rectum, 1993; 36:751–756.

20.

Kuhry

, Schwenk

, et al. Long-term outcome of laparoscopic surgery for colorectal cancer: A cochrane systematic review of randomized controlled trials. Cancer Treat Rev, 2008; 34:498–504.

21.

Buunen

, Veldkamp

, et al. Survival after laparoscopic surgery versus open surgery for colon cancer: Long term outcome of a randomized clinical trial. Lancet Oncol, 2009; 10:44–52.

22.

Klugsberg

, Haas

, et al. Current status of laparoscopic colonic surgery in Austria: A national survey. J Laparoendosc Adv Surg Tech, 2015; 25:976–981.

23.

Melich

, Jeong

, et al. Laparoscopic right hemicolectomy with complete mesocolic excision provides acceptable perioperative outcomes but is lengthy—Analysis of learning curves for a novice minimally invasive surgeon. Can J Surg, 2014; 57:331–336.

24.

Petereson

, Palazzi

, et al. The prevalence of laparoscopy and patient safety outcome: An analyisis of colorectal resections. Surg Endosc, 2014; 28:608–616.

25.

Alnasser

, Schneider

, et al. National disparities in laparoscopic colorectal procedures for colon cancer. Surg Endosc, 2014; 28:49–57.

26.

Xing-mao

, Hui-rong

, et al. Age, is it an obstacle for older surgeons to learn laparoscopic approach for colorectal cancer?. Med Oncol, 2013; 30:495.